Abstract: A system for automating evidence collection for software compliance is disclosed herein. The system can include a user computer device, a plurality of repositories, a backend computer system. The backend computer system can store a collection engine comprising a large language model (“LLM”) that, when executed by a processor, causes the backend computer system to extract metadata from a software architecture diagram and metadata from repository information, map the software architecture diagram to the at least one repository of the plurality of repositories based on the extracted metadata, generate a prompt to retrieve evidence from the plurality of repositories based on the mapping, and retrieve evidence relevant to the at least one requirement from the plurality of repositories based on the prompt.